How to find saliency map using ITTI's method?

11 次查看(过去 30 天)
Using spectral residual approach, i found the saliency map using the following matlab code:
%% Read image from file inImg = imread('60.jpg'); inImg = im2double(rgb2gray(inImg)); inImg = imresize(inImg, [64, 64], 'bilinear'); imshow(inImg);figure; %% Spectral Residual myFFT = fft2(inImg); myLogAmplitude = log(abs(myFFT)); myPhase = angle(myFFT); mySmooth = imfilter(myLogAmplitude, fspecial('average', 3), 'replicate'); mySpectralResidual = myLogAmplitude - mySmooth; saliencyMap = abs(ifft2(exp(mySpectralResidual + i*myPhase))).^2; %% After Effect saliencyMap = imfilter(saliencyMap, fspecial('disk', 3)); saliencyMap = mat2gray(saliencyMap); imshow(saliencyMap, []);figure; %% Object Map t = mean(mean(saliencyMap))*3; objectMap=saliencyMap>t; imshow(objectMap, []);
We compare our result with previous methods in the field, we also generate the saliency maps based on Itti’s well known theory [8] as a control set.The image is down-sampled to 320 × 240 for Itti’s method. For spectral residual method, each color channel is processed independently. In order to make a comparison, we must set either FAR or HR of the two methods equal. For instance, given the FAR of the spectral residual saliency maps, we can adjust the saliency map of Itti’s method S(x) by a parameter c: S'(x) = c*S(x) and use S'(x) instead of S(x) to compute FAR and HR in Eq.11 and Eq.12. Similarly, given the HR of Itti’s method, we linearly modulate the saliency maps generated by spectral residual.

回答(0 个)

类别

Help CenterFile Exchange 中查找有关 Biotech and Pharmaceutical 的更多信息

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by